<div dir="ltr"><div><font color="#000000">Dear all, <br></font></div><div><font color="#000000"><br></font></div><div><font color="#000000">We are looking for highly motivated students interested in doing an internship on movie understanding using large language models. The internship will be potentially continued with a joint PhD. <br></font></div><div><font color="#000000"><br></font></div><div><font color="#000000"><b>Description: </b>This project falls into the cutting-edge video understanding domain and aims to answer questions about movies. Movie Question Answering systems can provide a natural and intuitive way for users to interact with movies, making it easy and convenient to access information and perform movie-related tasks. Most computer vision approaches focus on answering specific questions about visual data, typically without long-term reasoning and without taking into consideration the particularities of the domain, such as jumps in time, cuts or intentions. Instead, given long-range video scenes, we will propose state-of-the-art methods designed to answer questions about movies, such as plot details, character information, and trivia by exploiting multimodal cues (video, audio, text). For this, we will exploit recent advances in Large Language Models to parse user queries and retrieve relevant visual information.</font></div><div><font color="#000000"><br></font></div><div><font color="#000000"><font face="arial, sans-serif"><b>Supervisors: </b></font><a href="http://vicky.kalogeiton.info/" target="_blank">Vicky Kalogeiton</a> (VISTA team, LIX  ́Ecole Polytechnique) and <a href="https://www.di.ens.fr/~laptev/" target="_blank">Ivan Laptev</a> (<a href="https://www.di.ens.fr/willow/" target="_blank">WILLOW project-tea</a>, Inria Paris)<font face="arial, sans-serif"> </font></font></div><div><font face="arial, sans-serif" color="#000000"><br></font></div><div><span lang="EN-US" style="font-family:arial,sans-serif;border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><font color="#000000"><font face="arial, sans-serif"><b>Location:</b> </font><span lang="EN-US" style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><a href="https://www.polytechnique.edu/en" style="border:0px;margin:0px;padding:0px;background-color:transparent" target="_blank">École Polytechnique</a></span></span></span><span lang="EN-US" style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px">, and <a href="https://www.inria.fr/fr/centre-inria-de-paris" target="_blank">Inria Paris</a>, Paris, France</span></span></span><br></font></span></span></div><div><font color="#000000"><br></font></div><div><font color="#000000"><b>Requirements: </b>We are looking for strongly motivated candidates with an interest in machine learning and computer vision. The project requires a strong background in applied mathematics and excellent programming skills (mostly in Python). <br></font></div><div><font color="#000000"><br></font></div><div><font color="#000000"><b>Application</b>: Potential applicants should send their complete CV and transcripts of grades to Vicky Kalogeiton at <a href="mailto:vicky.kalogeiton@polytechnique.edu" target="_blank">vicky.kalogeiton@polytechnique.edu</a> and Ivan Laptev at <a href="mailto:ivan.laptev@inria.fr" target="_blank">ivan.laptev@inria.fr</a>. </font>If needed, we may ask for two references. We particularly encourage applications from women and underrepresented groups in academia.<font color="#000000"><br></font></div><div><font color="#000000"><br></font></div><div><font color="#000000"><b>Application Deadline: </b>21 January 2023<br></font></div><div><font color="#000000"><br></font></div><div><font color="#000000"><span style="font-family:arial,sans-serif">For details, please see</span><span style="font-family:arial,sans-serif">: </span><a href="http://vicky.kalogeiton.info/assets/pdf/MovieQuestionAnswering2023.pdf" style="font-family:arial,sans-serif" target="_blank">PhD scholarship in Computer Vision for Movie Question Answering</a> or <span lang="EN-US" style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px"><span lang="EN-GB" style="border:0px;margin:0px;padding:0px"><span style="border:0px;margin:0px;padding:0px">contact </span></span></span></span></span></span><span style="font-family:arial,sans-serif">Vicky Kalogeiton </span><span style="font-family:arial,sans-serif">at </span><a href="mailto:vicky.kalogeiton@polytechnique.edu" style="background-color:transparent;font-family:arial,sans-serif;border:0px;margin:0px;padding:0px" target="_blank">vicky.kalogeiton@polytechnique.edu</a>. </font></div><font color="#888888"><div><br></div><div><br></div><div><br></div><div><font color="#000000">Vicky Kalogeiton</font></div></font></div>